The Ocean's Hidden Counterpoint: How Deep Mediterranean Water Flips Its Spin Over the Length of a Football Field
The Hidden Ocean Below the Mediterranean
Three kilometers beneath the sunlit surface of the Mediterranean Sea, where the pressure would crush an unprotected human lung in seconds, something unexpected is happening. Water is moving in slow, vast spirals—elliptical orbits that take hours to complete, driven by the distant, patient turning of the Earth itself. And in the spring of 2026, a Dutch oceanographer named Hans van Haren captured something about these movements that has eluded scientists for decades: evidence that the direction of rotation of these invisible water parcels flips back and forth over distances as short as fifty meters, governed by subtle variations in water density that most researchers had assumed were too small to matter.
The finding, published in July 2026 in the journal arXiv, concerns what oceanographers call "near-inertial" motions—oscillations in the deep ocean that rotate at roughly the same frequency as Earth's own spinning. These are not tides, not wind-driven currents, not the great conveyor-belt flows that carry heat around the planet. They are something more fundamental: the ocean's response to its own rotation, the way a cup of coffee sloshes differently on a spinning carousel than on a stationary table. Understanding them matters because they sit at the intersection of two of the most chaotic, least predictable regimes in physical oceanography—internal waves and sub-mesoscale eddies—and because they may help explain how energy cascades from the large-scale ocean circulation down to the tiny turbulent swirls where it finally dissipates as heat.
Van Haren's team deployed a complex mooring system in water more than 25,000 meters deep, scattered with instruments every few tens of meters like a vertical forest of sensors. They let it sit for weeks, listening to the deep ocean's whispers. What they found was that water parcels separated by just fifty meters could be moving in opposite rotational directions at the same time—a kind of oceanic counterpoint that textbooks had predicted but never directly observed. And the conductor governing this counterpoint appears to be something as seemingly trivial as how much the water's density changes with depth: a quantity so subtle that it hovers near the threshold of what oceanographers can even measure.
The Science
The study emerged from a decades-long research program focused on understanding the deep Mediterranean, one of the most peculiar bodies of water on Earth. Unlike the Atlantic or Pacific, the Mediterranean is almost entirely isolated, connected to the global ocean only through the narrow Strait of Gibraltar. Water evaporates furiously from its surface, making it saltier and denser than the Atlantic inflow, which sinks and flows eastward along the bottom before welling up again near the Levant. This circulation makes the Mediterranean a natural laboratory for studying ocean physics without the complications of polar inputs or equatorial heating.
What makes the deep Mediterranean especially valuable for this kind of research is its stratification—or rather, its lack of it. In most of the ocean, the density of seawater increases with depth as cold, salty water sinks below warmer, fresher water. This density gradient acts like a spring, restoring water parcels that are pushed up or down and allowing internal waves to propagate. The "strength" of this spring is measured by the buoyancy frequency, denoted N: a higher N means stronger stratification, faster wave propagation, and more energetic internal wave activity. In the open ocean, N is typically ten to a hundred times larger than the inertial frequency f, the frequency at which water parcels oscillate due to Earth's rotation.
In the deep Mediterranean, van Haren and colleagues had previously discovered something unusual: the buoyancy frequency is nearly equal to the inertial frequency. In this regime, the standard equations governing internal waves break down, and a more complex theory—the "non-traditional" inertio-gravity wave theory—takes over. The difference matters. Under normal conditions, near-inertial waves have a predictable clockwise (anticyclonic) polarization in the Northern Hemisphere, like water circling a low-pressure system. But in the Mediterranean deep, where N ≈ f, the polarization becomes unstable, flipping between clockwise and counterclockwise (cyclonic) depending on local conditions.
The instruments van Haren's team deployed were current meters—devices that measure water velocity using acoustic pulses or mechanical sensors—strung along a cable anchored to the seafloor and held upright by buoys. But this was no ordinary mooring. To capture the fine-scale polarization variations the researchers were hunting, they needed instruments spaced every few tens of meters across a vertical array, with synchronized clocks accurate to milliseconds. The resulting setup, which the team calls a "complex mooring system," recorded horizontal velocity components at 15-25 meter intervals throughout the water column, from near the bottom to within a few hundred meters of the surface.
The data collection ran for several weeks, capturing a continuous time series of deep-ocean velocities. The researchers then analyzed these time series using spectral methods—mathematical techniques that decompose a time series into its frequency components, revealing the dominant periodicities in the flow. Crucially, they also computed the "ellipse parameters" of the near-inertial oscillations: the orientation, eccentricity, and phase of each elliptical orbit, which together determine the polarization direction.
One of the key challenges was distinguishing between true polarization changes—switches in the direction of rotation of the elliptical motion—and artifacts caused by sampling limitations or instrumental noise. Van Haren's team addressed this by examining multiple frequency bands within the near-inertial range, by comparing results from adjacent instruments, and by using statistical tests to assess the significance of detected polarization flips. The fact that the alternation was observed "varyingly over 50-m distances" suggests it is a real spatial structure, not random noise.
What They Found
The most striking result is the magnitude of the horizontal velocity differences observed over small spatial scales. Water parcels separated by just fifty meters were moving relative to each other at speeds up to 0.02 meters per second—roughly four centimeters per second, or about one-tenth of a typical walking pace. That may sound slow, but in the deep ocean, where total current speeds rarely exceed 0.07 meters per second (about two and a half kilometers per hour), it represents a substantial fraction of the total kinetic energy. More importantly, this shear—the gradient of horizontal velocity with distance—caused the relative vorticity to approach the inertial frequency f. In plain terms, the water was spinning locally nearly as fast as it oscillates in response to Earth's rotation. This is a regime that standard oceanographic theory treats as a boundary condition, not a central feature of the flow.
The polarization analysis revealed a pattern that had been theoretically predicted but never directly observed at this fidelity. Under near-homogeneous conditions—when the water column was well-mixed and the buoyancy frequency dropped to very low values—near-inertial motions exhibited anticyclonic polarization. This is the "traditional" behavior, consistent with standard inertio-gravity wave theory and with the handedness of Earth's rotation in the Northern Hemisphere. But when the water became more stratified, the polarization flipped to cyclonic, reversing the direction of the elliptical orbit. Crucially, this flip did not occur uniformly across the water column. It varied over distances of fifty meters or less, suggesting that the stratification itself was patchy at these scales—a finding that contradicts the assumption of smooth, slowly varying backgrounds that underlies most theoretical treatments.
Perhaps the most significant discovery is that the cyclonic polarization under stratified conditions was only observed when "relatively strong turbulent convection" was simultaneously present. In other words, the theory that predicts polarization flips requires not just stratification but also active mixing—turbulent eddies that reduce stratification locally by dragging denser water downward or lighter water upward. This is a subtle and important point: the internal wave dynamics are not governed by the background stratification alone but by the interaction between stratification and turbulence, a coupling that standard models struggle to capture.
The data also revealed a temporal pattern in the polarization states. The alternation between anticyclonic and cyclonic polarization was not random; it appeared to be organized in episodes, with each polarization state persisting for days to weeks before switching. This temporal structure likely reflects the passage of sub-mesoscale eddies—coherent vortices with horizontal scales of tens of kilometers and lifetimes of weeks—through the mooring array. These eddies would modulate the local stratification as they passed, switching the internal wave polarization on and off like a light controlled by a dimmer.
Why This Changes Things
The implications of this study extend far beyond the Mediterranean. Near-inertial motions are a universal feature of the ocean, generated by winds, tides, and instabilities of the mean circulation. They carry energy from the surface into the deep, where they break, mix, and ultimately dissipate. Understanding their polarization—and how that polarization depends on local stratification—is therefore essential for quantifying the ocean's energy budget and its role in climate.
The standard picture of near-inertial waves treats polarization as a fixed property, determined by latitude and wave frequency. In this picture, waves in the Northern Hemisphere always rotate clockwise, and any deviation is a sign of contamination by other processes. Van Haren's results challenge this picture fundamentally. In the deep Mediterranean, where buoyancy frequency and inertial frequency are comparable, polarization is not fixed but flexible, flipping in response to subtle variations in stratification and turbulence. This suggests that the standard theory, while correct in the "traditional" regime where N >> f, is incomplete for the large swaths of the ocean where the two frequencies are comparable—which includes much of the deep ocean and, importantly, the ocean's interior in the tropics and subtropics.
The finding also has implications for the study of internal wave-wave interactions. When near-inertial waves have a fixed polarization, certain nonlinear interactions are forbidden by symmetry. When polarization is flexible, these interactions become possible, potentially opening new channels for energy transfer. This could matter for the global distribution of ocean mixing: if the Mediterranean regime is common in the deep ocean, it would imply a richer and more efficient cascade of energy from large scales to small turbulent scales, where it is finally dissipated as heat.
Perhaps most provocatively, the study suggests that sub-mesoscale eddies—the thousand-kilometer-per-day vortices that are notoriously difficult to observe directly—may have a larger influence on internal wave dynamics than previously thought. By modulating local stratification over horizontal scales of tens of kilometers, these eddies could act as switches, turning internal wave interactions on and off as they pass through a given region. This would connect two fields that have developed largely in parallel: the study of internal waves and the study of sub-mesoscale dynamics.
What's Next
Several questions emerge directly from this study. First, how common is the Mediterranean regime—where buoyancy frequency is on the order of the inertial frequency—across the global ocean? Existing climatologies suggest it may be widespread in the deep ocean and in semi-enclosed basins, but direct observations like van Haren's are rare. A coordinated global program to deploy similar high-resolution mooring arrays in different ocean basins would help answer this question.
Second, what are the consequences of flexible polarization for the ocean's energy budget? The study shows that polarization flips are associated with increased horizontal velocity differences, which in turn imply larger shear and potentially more vigorous mixing. But quantifying this mixing will require combining the velocity observations with direct microstructure measurements—fine-scale measurements of the rate at which kinetic energy is converted to heat by viscous dissipation.
Third, how do the sub-mesoscale eddies that seem to organize the polarization states form and evolve? The study was not designed to resolve eddy dynamics directly; it captured the signature of eddies passing through the array, but not their detailed structure. Future studies could combine fixed moorings like van Haren's with autonomous underwater vehicles or satellite altimetry to build a three-dimensional picture of how eddies modulate internal wave polarization.
Finally, the theoretical framework itself deserves scrutiny. The "non-traditional" inertio-gravity wave theory that predicts polarization flips is decades old, but it has rarely been tested against direct observations. Van Haren's data provide the most direct test to date, and the agreement is striking—but not perfect. The theory predicts cyclonic polarization under stratified conditions; the observations show it, but only in the presence of turbulent convection. This suggests that even the non-traditional theory is incomplete, and that a new synthesis—including a more complete treatment of turbulent mixing—may be needed.
For the broader oceanographic community, the study is a reminder that the deep ocean is not a calm, uniform reservoir but a dynamic, finely structured environment where processes operating at scales of meters can influence phenomena at scales of kilometers. It is a call to look more carefully, measure more precisely, and resist the temptation to assume that the ocean's interior is boring just because it is dark and cold and far from human experience. In the deep Mediterranean, three kilometers down, the water is spinning in circles that reverse direction fifty meters apart—and that, it turns out, tells us something important about how the ocean works.
Hans van Haren is a senior researcher at the Royal Netherlands Institute for Sea Research (NIOZ). His work focuses on high-resolution observations of internal waves, turbulence, and mixing in the deep ocean. The data described in this paper were collected as part of the Deep-Mediterranean Mixing experiment.
Technical Appendix: Understanding the Key Concepts
Inertial Frequency
The inertial frequency, denoted f, is the frequency at which a water parcel in the ocean oscillates when displaced horizontally and allowed to move freely under the influence of Earth's rotation. It depends only on latitude: at 45° North or South, f corresponds to a period of about 17 hours. Near-inertial motions are those whose frequency is close to f; they are generated by winds, storms, and other disturbances that impart a rotational impulse to the upper ocean.
Relative Vorticity
Vorticity is a measure of local rotation in a fluid. "Relative" vorticity refers to the rotation of the fluid relative to the fixed Earth, as opposed to the planetary vorticity associated with Earth's rotation. When relative vorticity equals the inertial frequency f, the local Rossby number—an important dimensionless parameter—approaches unity, indicating that the flow is in a regime where rotation and inertia are equally important.
Buoyancy Frequency
The buoyancy frequency, denoted N, is the frequency at which a vertically displaced water parcel oscillates under the restoring force of stratification. It is defined as the square root of the squared buoyancy gradient: $N^2 = -\frac{g}{\rho_0}\frac{d\rho}{dz}$, where g is gravitational acceleration, $\rho_0$ is a reference density, and $d\rho/dz$ is the vertical density gradient. A higher N means stronger stratification and faster internal wave propagation. In the Mediterranean deep, $N$ is unusually low, making it comparable to f.
Polarization
In the context of internal waves, polarization refers to the orientation and direction of traversal of the elliptical orbit traced by a water parcel as it oscillates. In the Northern Hemisphere, traditional near-inertial waves rotate clockwise (anticyclonically) when viewed from above. Non-traditional theory predicts that under certain conditions—specifically, when $N$ is comparable to f—the rotation can reverse, becoming counterclockwise (cyclonic). The switch is controlled by the local balance between stratification and Earth's rotation.
Non-traditional Inertio-gravity Wave Theory
Standard internal wave theory includes the vertical component of the Coriolis force (associated with Earth's rotation) but neglects the horizontal component. This "traditional" approximation is valid when the horizontal wavelength is much smaller than the internal Rossby radius of deformation—a condition met in most of the upper ocean but violated in the deep ocean and in regions of weak stratification. Non-traditional theory retains both Coriolis components, leading to modified wave properties, including the possibility of cyclonic polarization under stratified conditions. The present study provides direct observational confirmation of this theoretical prediction.